Abadh Kishor Jhaa,
Shyam Kishorb and
Nidhi Jain*a
aDepartment of Chemistry, Indian Institute of Technology, New Delhi-110016, India. E-mail: njain@chemistry.iitd.ac.in; Fax: +91 11 26581102; Tel: +91 11 26591562
bDepartment of Chemistry, J.V. College, Uttar Pradesh-250611, India
First published on 10th June 2015
Palladium catalyzed cross-coupling with aryl chlorides has always been a challenge and does not take place without the assistance of bulky phosphine ligands, NHCs or other additives. In this work, we present an unprecedented phosphine-free coupling of cheap and less reactive substrates like aryl chlorides with 1-(bromoethyl)benzene as a nucleophilic coupling partner at low Pd concentrations, under microwave conditions to access highly regio- and stereospecific trans-stilbene derivatives. The reaction is believed to proceed via a Pd–carbene intermediate generated in situ under MW conditions from 1-(bromoethyl)benzene. Another key requirement in the reaction is the presence of styrene; as in its absence, Pd catalysed coupling of halogenated ethyl benzene with chloroarenes does not take place at all. DFT studies have also been carried out to obtain a thermodynamic insight into the proposed mechanism.
Since its discovery, the Mizoroki–Heck reaction which involves coupling of aryl iodides, bromides, triflates or diazonium salts with styrene in the presence of palladium catalyst and stoichiometric amount of base remains the best method for preparing stilbene derivatives.7 While the method has been a subject of extensive research and taken up by chemical industry, the process economics remains a major concern.8,9 It demands that the reaction should occur at low catalyst loadings whilst providing the product in high yield, follow a ligand-free protocol, and make use of readily available and cheap starting materials like organic chlorides instead of bromides and iodides. The limitation with chloroarenes, however, is that unlike aryl iodides and bromides, they are unreactive under the standard reaction conditions.10 Catalytic systems especially designed for cross-coupling with aryl chlorides include air-sensitive, bulky and sometimes toxic phosphine ligands.11–16 An additional problem encountered in stilbene synthesis via Heck coupling is that the coupling partners, styrenes have to be used in excess owing to a possible polymerization under usual high temperature conditions of the reaction.17
A solution to this problem has been offered in two former reports which demonstrate a Heck coupling by generating olefin from alkyl halides or alcohols in situ.18 While a dehydrohalogenative as well as a dehydrative Heck olefination with aryl iodides and bromides has been reported, its success with the less reactive and challenging aryl chlorides is less known.19 In our efforts directed towards lifting these limitations and developing an economically useful C–C coupling involving a ligand-free C–Cl activation; we investigated the reaction of aryl chlorides with secondary alkyl bromides as coupling partners under microwave conditions. The reaction believed to proceed via Pd–carbene complex yields exclusively the β-arylated product. This is in contrast to the Pd-catalyzed cross-coupling of hydrazones with aryl halides (also involving a Pd–carbene complex) which affords the α-arylated product selectively (Scheme 1E).20
:
3 as determined by HPLC (ESI Fig. S1†) was taken in DMF. To it, chlorobenzene (3.0 equiv.), followed by KOH (2.0 equiv.) in water was added. Finally, Pd(OAc)2 (1.0 mol%) was added to the reaction mixture and it was irradiated in microwave for one hour at 140 °C. We were delighted to find that the desired product trans-stilbene (3a) was formed in 40% yield (Table 1, entry 2) along with unreacted 1a and styrene while no traces of 1,1-diphenylethene (3a′) were seen (Scheme 2).19 The product showed a singlet at 7.18 ppm in 1H NMR for two olefinic protons indicating it to be a symmetrical alkene. Peaks in the aromatic region between 7.28–7.59 ppm for 10 protons confirmed the presence of two phenyl rings and λmax at 296 nm in UV suggest it to be the trans- isomer of stilbene (ESI Fig. S2†).
| Entrya | Substrate | Styrene (equiv.) | Pd(OAc)2 (mol%) | Yield (%; HPLC) |
|---|---|---|---|---|
a Reaction conditions: substrate (1a–e) (1.0 equiv., 0.5 mmol), styrene, chlorobenzene (2a; 3.0 equiv.; 2.0 mmol), KOH (3.0 equiv.), Pd(OAc)2, DMF : H2O (1 : 1), MW, 140 °C, 50 min.b KOH (2.0 equiv.).c No styrene added.d PPh3 added in place of styrene.e Pd(CH3CN)2Cl2 instead of Pd(OAc)2. |
||||
| 1 | ![]() |
— | 0.5 | 0 |
| 2 | ![]() |
0.3 | 1.0 | 90, 40b |
| 3 | ![]() |
0.3 | 0.5 | 93, 0c, 0d, 63e |
| 4 | ![]() |
0.4 | 0.5 | 60 |
| 5 | ![]() |
0.2 | 0.5 | 70 |
| 6 | ![]() |
0.3 | 0.25 | 66 |
| 7 | ![]() |
0.3 | 0.5 | 80, 0c |
| 8 | ![]() |
0.3 | 0.5 | 67, 0c |
| 9 | ![]() |
0.3 | 0.5 | 45, 0c |
| 10 | ![]() |
0.3 | 0.5 | 33, 0c |
In a surprising revelation, we discovered that when the same reaction was carried out with pure 1a without any traces of styrene, the desired product 3a was not formed at all (Table 1, entry 3). Also, interestingly as expected, another reaction with styrene as the starting substrate failed to yield stilbene under the employed reaction conditions, and was recovered as such (Table 1, entry 1) after the reaction. These observations were quite unique and important from the perspective of ligand-free C–C bond formation with aryl chlorides as coupling partners, and motivated us to explore the reaction in greater details. We were excited to find that lowering the Pd concentration from 1 mol% to 0.5 mol% and increasing KOH from 2.0 equiv. to 3.0 equiv. increased the yield of 3a drastically from 40 to 93% (Table 1, entry 2 and 3). We surmised that the reason for an increased yield might be vested in particle size reduction of active Pd catalyst on lowering Pd concentration. With 0.5 mol% Pd loading, DLS demonstrated the existence of Pd-colloids with an average diameter of 17 nm (ESI Fig. S3†). It is believed that low Pd concentrations suppress the formation of Pd-black keeping all the metal available for catalysis. Previously, de Vries and co-workers demonstrated a ligand-free Heck olefination of styrene with activated and unactivated aryl bromides and aryl iodides, using 0.05 mol% of Pd(OAc)2 with NaOAc as base in NMP at 135 °C.21 Their system, however failed completely with aryl chlorides as substrates. Through this work, however, we report for the first time a phosphine-free coupling of aryl chlorides with secondary alkyl bromide instead of styrene as coupling partner at low Pd loading. On changing the equivalents of styrene to 0.4 or 0.2, the yield of 3a dropped to 60 and 70% respectively (Table 1, entry 4, 5) suggesting that 30% styrene in the reaction mixture was optimum for high product yield.
To ascertain the role of styrene in the reaction, it was replaced by Ph3P (Table 1, entry 3). However, under these conditions, compound 3a was not formed at all suggesting that styrene as an additive was not facilitating the reaction simply by acting as a ligand, but had some hitherto unknown effect on kinetics of the reaction. Further, reactions of chlorobenzene with primary halide, 2-(bromoethyl)benzene (1b) under the same reaction conditions afforded 3a in 80% yield (Table 1, entry 7). This prompted us to explore the effect of substrates in controlling coupling with chlorobenzene. For this, iodo- (1c), chloro- (1d) and hydroxyl (1e) analogues of 1a were prepared.
As was seen with 1a, in the absence of styrene, 3a was not formed at all in all the three cases (Table 1; entry 8, 9, 10). While some unidentified polar impurities were obtained with 1c and 1e, the chloro- analogue (1d) was recovered as such after the reaction. However, on addition of 0.3 equivalent of styrene to (1c–e), 3a was formed in 33–67% yields (Table 1; entry 8, 9, 10). All these results pointed out towards the fact that the reaction as anticipated early on did not follow a usual dehydrohalogenative Heck pathway involving an intermediate styrene. In fact, since all the substrates (1a–e) yielded 3a as the major product, it indicated that all these reactions proceeded via a common intermediate which underwent facile coupling with chlorobenzene in contrast to previous reports where β-elimination mediated styrene formation was the preferred pathway.18,19 Furthermore, it was apparent that presence of styrene (30 mol%) was inevitable for driving the reaction towards stilbene formation.
Since, the highest yield of coupled product 3a was obtained from compound 1a, further optimization to analyse the effect of solvent, temperature, time, amount of base, nature of base, and amount of chlorobenzene was carried out with 1a as the substrate (Table 2).
| Entrya | Solvent | Base | Yield (%; HPLC) | Time (min) |
|---|---|---|---|---|
| a Reaction conditions: 1-(bromoethyl)benzene, 1a (1.0 equiv.), styrene (0.3 equiv.), chlorobenzene (3.0 equiv.), base (3.0 equiv.), Pd (OAc)2 (0.5 mol%), MW, 140 °C.b KOH (2.0 equiv.).c Chlorobenzene (2.0 equiv.).d Reaction done at 120 °C.e Reaction done at 160 °C.f Conventional heating at 140 °C for 24 h. | ||||
| 1 | DMF | KOH | 85 | 50 |
| 2 | CH3CN | KOH | 0 | 50 |
| 3 | H2O | KOH | 54 | 50 |
| 4 | DMF/H2O | KOH | 93, 38b, 21c | 50 |
| 5 | CH3CN/H2O | KOH | 0 | 50 |
| 6 | Acetone/H2O | KOH | 56 | 50 |
| 7 | THF/H2O | KOH | 6 | 50 |
| 8 | DMSO | KOH | 0 | 50 |
| 9 | DMSO/H2O | KOH | 0 | 50 |
| 10 | DMF/H2O | Et3N | 6 | 50 |
| 11 | DMF/H2O | K2CO3 | 66 | 50 |
| 12 | DMF/H2O | — | 0 | 50 |
| 13 | DMF/H2O | [bmim]OH | 48 | 50 |
| 14 | DMF/H2O | KOH | 25d | 50 |
| 15 | DMF/H2O | KOH | 90e | 50 |
| 16 | DMF/H2O | KOH | 90 | 60 |
| 17 | DMF/H2O | KOH | 63 | 40 |
| 18 | DMF/H2O | KOH | 37, 0f | 30 |
A series of polar solvents including DMF, CH3CN, DMSO, water, DMF/H2O, CH3CN/H2O, DMSO/H2O, THF/H2O and acetone/H2O were screened. The results indicated that with 3.0 equivalent of KOH, highest yield of 3a (93%) (Table 2, entry 4) was obtained in DMF/H2O (1
:
1) in 50 minutes at 140 °C. On the contrary, in DMF and water alone as solvents, yield dropped to 85% and 54% respectively (Table 2, entry 1 and 3) owing to insolubility of base in one case and formation of a biphasic reaction mixture in the other. CH3CN, DMSO and their aqueous solutions turned out to be poor choice of solvent and failed to yield any product (Table 2, entry 2, 5, 8, 9). Screening of various bases such as KOH, K2CO3, [bmim]OH, and triethylamine showed that KOH worked best under the employed conditions (Table 2, entry 4, 10, 11, 13). Lowering the amount of KOH to 2.0 equivalents lowered the yield to 38% (Table 2, entry 4). Further, temperature variation in the range 120–160 °C demonstrated that 140 °C was ideal for the reaction. At 120 °C, reaction did not go to completion and 3a was obtained in only 25% yield (Table 2, entry 14) while at 160 °C, yield was 90% which was almost the same as at 140 °C (Table 2, entry 15). Next, the reaction was monitored with respect to varied reaction times ranging from 30–60 minutes under MW irradiation and 24 hours under conventional heating. While the yield dropped significantly on decreasing the reaction time in MW (Table 2, entry 17, 18), to our astonishment, no trace of product was seen even after 24 hours under conventional heating. This observation emphatically highlighted the critical role of MW in producing an intermediate necessary for the progress of the reaction and its coupling with aryl chlorides. The best reaction conditions obtained after extensive optimization studies clearly demonstrated the high activity and efficiency of our catalytic system compared to any previous literature report on stilbene formation via cross-coupling with aryl chlorides.22
Under optimized conditions, we furthered our investigation into the scope of aryl chlorides as substrates for synthesis of substituted trans-stilbenes, and the results are summarized in Table 3. Both activated as well as inactivated aryl chlorides afforded moderate to high yield of products (70–90%), except for 4-chloromethoxy benzene (2i) which gave 1-methoxy-4-styrylbenzene (3i) in 20% yield. With chloro derivatives of pyridine (2j) and thiophene (2k), corresponding heterocyclic stilbene derivatives 3j and 3k were obtained in moderate yields. The generality of the reaction was further explored with 2-(1-bromoethyl)naphthalene (1a′) as the substrate. Under similar reaction conditions, substrate 1a′ after reaction with aryl chlorides afforded corresponding trans-olefins (3m–o) in moderate yields. All these studies clearly stated that for a phosphine-free C–C coupling of 1a or 1a′ with chloroarenes at low Pd loading, 0.3 equivalent of styrene and MW conditions were indispensable.
| Entrya | Aryl halide (2) | Product (3) | Yieldb (%) |
|---|---|---|---|
a Reaction conditions: 1a (1.0 equiv.), aryl chloride (3.0 equiv.), Pd(OAc)2 (0.005 equiv., 0.5 mol%), KOH (3.0 equiv.), styrene (0.3 equiv.), DMF/water (1 : 1), 50 minutes under MW.b Isolated yield with respect to 1a/1a′.c 1.0 equiv. of TEMPO was added.d Reaction was carried out with 1a′ (1.0 equiv.) and vinyl naphthalene (0.3 equiv.) in place of 1a and styrene. |
|||
| 1 | ![]() |
![]() |
86 (0)c |
| 2 | ![]() |
![]() |
84 |
| 3 | ![]() |
![]() |
90 |
| 4 | ![]() |
![]() |
85 |
| 5 | ![]() |
![]() |
82 |
| 6 | ![]() |
![]() |
80 |
| 7 | ![]() |
![]() |
74 |
| 8 | ![]() |
![]() |
70 |
| 9 | ![]() |
![]() |
20 |
| 10 | ![]() |
![]() |
74 |
| 11 | ![]() |
![]() |
70 |
| 12 | ![]() |
![]() |
80 |
| 13 | 2a | ![]() |
84d |
| 14 | 2k | ![]() |
74d |
| 15 | 2l | ![]() |
76d |
![]() | ||
| Fig. 2 300 MHz 1H NMR spectrum of reaction mixture taken after 5 minutes. Lower spectrum (blue) in CDCl3 and upper spectrum (red) showing D2O exchange. | ||
The emergence of these new signals pointed out towards formation of multiple species and intermediates in the reaction. The chemical shifts at 4.85(q) and 1.47(d) ppm were assigned to the CH and CH3 protons of 1-(hydroxyethyl)benzene (1e) which might have formed from 1a as a result of microwave assisted C–Br bond cleavage to yield benzyl radical (I) which upon attack from water in the medium afford 1e as shown in Fig. 3 (path I).
The disappearance of hydroxyl proton at 3.2 ppm on D2O exchange confirmed the formation of 1e during the reaction (Fig. 2). Apparently, radical intermediate I undergoes loss of hydrogen radical to yield a Pd–carbene complex (II) which is suggested by the methyl singlet at 2.12 ppm (Fig. 2). The chemical shift at 3.21(q) ppm pointed out towards a significantly shielded benzylic proton which was originally at 5.02 ppm in 1a. This shielding was contemplated to be a consequence of the coordination of benzylic carbon to Pd.
It was envisaged that the reaction was also going through an alternate path of oxidative addition of 1a to Pd(0) to yield intermediate VI which gave a doublet at 1.34 ppm and a quartet at 3.21 ppm in the 1H NMR spectrum (Fig. 1 and 2). Increasing the reaction time to 10 and 20 minutes did not bring about any noticeable change in the peaks of 1H NMR spectra. The presence of multiple species that is 1-(hydroxyethyl)benzene (1e), Pd–carbene complex II and the oxidative addition intermediate VI was also evident from the 13C NMR spectrum of the reaction mixture acquired at 10 min (see ESI,† spectrum 60). The benzylic carbon of 1e and VI appeared at 70.1 and 66.0 ppm respectively consistent with the shielding behavior of the corresponding protons in the 1H NMR spectra. Similarly, the methyl group in 1e, VI, and II appeared at 25.3, 20.2, and 43.1 ppm respectively. Further, DEPT and COSY data at 10 min. supported the co-existence of these species in the reaction mixture (see ESI Fig. S4–S6†). LC-MS study of the reaction mixture at 1, 5, and 10 minutes showed evidence for the existence of Pd–carbene complex II (see ESI Fig. S7–S9†). The presence of Pd in the complex was also confirmed from EDAX spectra of the reaction mixture acquired after 5 minutes (see ESI Fig. S10†). After 50 minutes, formation of trans-stilbene was evident from the disappearance of methyl singlet at 2.1 ppm and appearance of new peaks in the aromatic region. Further, X-ray23 structure of compound 3c confirmed the trans-geometry (Fig. 4; see ESI Fig. S11 and Table S1† for details).
The proposed mechanism as shown in Fig. 3, supported largely by the NMR and mass spectral evidences was verified further by DFT studies (Fig. 5). The aim was to identify intermediates which are thermodynamically stable and to compute free energy of overall reaction. In a way, we attempted to provide conformity to the formation of the species contemplated on the basis of the NMR spectra. The catalytic cycle starts with oxidative addition of Pd(0) to compound 1a to yield an intermediate VI which is predicted to be quite stable. The relative free energy (G°rel) of intermediate VI in binary solvent is −163.72 kJ mol−1. This is also justified by its appearance and existence till 40 minutes in the 1H NMR spectrum (1–40 min.). Dehydrohalogenation of VI yields the Pd–carbene monomer II (G°rel = 8.69 kJ mol−1). Complexation with π-donor ligands like styrene/vinyl naphthalene is predicted to provide the resting state (transient intermediate) to the Pd–carbene (see ESI Fig. S12†). Similar formation of ligand stabilized Pd–carbene complexes is already known in literature19 and has also been theoretically verified by DFT calculations.19 The G°rel of ligand stabilized Pd–carbene complex predicts that in the presence of one ligand, the electrosteric stabilization is maximum. At the B3LYP level of theory, Pd–carbene when attached to one ligand has a lower relative free energy than with two ligands (see ESI Fig. S12†). These transient intermediates are expected to promote C–Cl bond activation in chloroarenes resulting in the formation of intermediate III (G°rel = −10.39 kJ mol−1). Aryl migration in III can lead to either IV (G°rel = −169.75 kJ mol−1) or IVa (G°rel = −156.11 kJ mol−1). Because of less steric hindrance, complex IV shows greater stability than IVa by 13.64 kJ mol−1. This is in agreement with the experimentally isolated product, trans-stilbene (3a) obtained from β-elimination of IV in contrast to 1,1-diphenyl ethane (3a′) which was not isolated. The free energy of the overall reaction is −42. 78 kJ mol−1.
![]() | ||
| Fig. 5 Potential energy surface of mechanistic pathway postulated in Fig. 3. Relative Gibbs Free Energy (G°rel) given in kJ mole−1. G°rel calculated w.r.t. 1-(Bromoethyl)benzene 1a (G°1a = 0.0) inbinary solvent DMF–H2O. For structure clarification, see ESI Table S2–S19.† | ||
Footnote |
| † Electronic supplementary information (ESI) available: Detailed experimental procedures, copies of 1H NMR, 13C NMR and mass spectra for all the new compounds, X-ray crystallographic data for 3c, and computational data related to the optimized geometry of intermediates and transition states. HPLC, UV-Vis, EDAX and LC-MS data available. CCDC 1028616. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c5ra06107e |
| This journal is © The Royal Society of Chemistry 2015 |